Modelling soil-regolith thickness in complex weathered landscapes of the central Mt Lofty Ranges, South Australia J. Wilford M. Thomas

• Published in: Digital Soil Assessments and Beyond pp. 69 - 76
• Format: Book Chapter
• Language: English
• Published: CRC Press 2012
• Subjects: SRTM DEM; Weathered Bedrock; Traditional Soil Mapping; Geoscience Australia; Soil Survey; Airborne Gamma Ray; Highly Weathered; Soil Mapping; Regression Kriging; Environmental Covariates; Original Bedrock Fabric; Homogeneous Sub-regions; Paddock Scale; Drainage Enforcement; Muscovite; Detailed Soil Mapping; Kangaroo Island; Kanmantoo Group; Soil Landscape Models; Weathered Zone; AB Profile; Study Area Landscape; Fault Scarps; Regolith Cover
• DOI: 10.1201/b12728-18

Thickness and characteristics of soil-regolith profoundly govern groundwater interactions and subsoil water movement, water storage and nutrient availability. As such, the soil-regolith has important bearing on land use, and the viability of land-based industries dependant on rooting depth, e.g., agriculture and forestry. In addition, from a mineral exploration perspective, the surface geochemical expression of buried deposits is often intrinsically linked to the nature and thickness1 INTRODUCTIONA soil profile comprises an A and B horizon. This profile with AB features horizons that have undergone strong pedogenesis by chemical, physical and biological action to such a degree that the material retains none or little of the fabric of the rock below, i.e., the presumed parent material in an in situ soil profile. The C horizon is a mineral layer below the AB profile that retains at least some of the rock fabric due to variable degrees of in situ weathering. The composition and fabric of the C horizon can range from almost complete mineral alteration, to secondary minerals (e.g., Fe-oxides and clays) with only the most resistant minerals (e.g., quartz) being retained, to moderately weathered material which retains much of the primary mineralogy and fabric of the bedrock. Deeper still is the R layer, which consists of a continuous mass of rock that may have undergone minor weathering along fractures and bedding/cleavages. These horizons broadly equate with completely, highly, moderately and slightly weathered bedrock zonesof the regolith cover. Given this importance, therefore, it is surprising that there is a dearth of specific depth information incorporated in most soil-regolith mapping. There is a mature tradition in traditional soil mapping of relying on conceptualisation of soil-landscape models. These models are based on the interaction of multiple soil forming factors, including relief, biology, climate, parent material and age (Jenny, 1941). Soil-landscape models ranging from catchment to paddock scale (e.g., 1:100,000-1:10,000 scale) typically bear the strongest overprint of relief, with the other factors influencing to lesser and varying degrees. One of the key factors influencing soil-regolith thickness is landscape age and the associated processes of regolith formation and removal through time. Most soils maps and associated descriptions focus on the upper A and B horizons of the weathered zone or, in the case of deep regolith, the last pedological overprint on invariably a much older weathering profile. Regolith-landform maps are typically under-pinned by a stronger focus on landscape age (often conceptual) through an understanding or reconstruction of landscape evolution through time. However, both mapping approaches are limited by a scarcity of accurate site depth information, typically beyond a couple of meters. Drill logs, e.g., as used in mining, stratigraphic and hydrological work, can provide useful information. However descriptions from these logs are typically difficult to interpret or translate into depth estimates due to a lack of standardised terminology or coring methods. Despite detailed soil mapping and drilling throughout the Mt Lofty Ranges there was limited site information on soil-regolith thickness for modelling soil-regolith depth. This lack of deeper subsurface information restricted our depth prediction to the upper part of the weathered zone but allowed us to incorporate observations from road cutting and gully exposures to populate and constrain the model.